1. Field of the Art
The disclosure relates to the field of automatic vector registration to linear features in remotely-sensed data.
2. Discussion of the State of the Art
In the art of vector registration, ROADTRACKER™ and similar tools enable automatic registration of existing two-dimensional linear feature vectors, possibly of poor geospatial accuracy, to the two-dimensional centerlines of corresponding linear features (e.g., roads, trails, single-line drainage, boundaries, etc.) in remotely-sensed imagery. This capability is frequently called (two-dimensional) vector-to-image registration. The resulting registered vectors maintain the interconnection topology of the original vectors. The discrepancy between the original vectors and their respective linear feature centerlines may be due, for example, to geometric over-generalization in the original vector extraction, or mistakes in the original extraction, or may be due to the original extraction being performed against other data or imagery that was itself not well-registered to the current imagery. The automatic registration tools allow for a visual comparison of the before-and-after vectors against the background context of the current imagery.
What is needed is a tool to enable automatic registration of existing three-dimensional linear feature vectors, possibly of poor geospatial accuracy, to the three-dimensional centerlines of corresponding three-dimensional linear features in remotely-sensed data. The resulting registered vectors should maintain the interconnection topology of the original vectors. The tool should allow for visual comparison of the before-and-after vectors against the background context of the remotely-sensed data.